TY - GEN
T1 - Graphical models of multivariate volatility
AU - Rea, Alethea
AU - Reale, Marco
AU - Scarrott, Carl
PY - 2007
Y1 - 2007
N2 - In order to understand volatility transmission between financial assets a multivariate model is essential. This paper looks at using graphical modelling to study volatility transmission. Graphical modelling is a technique that objectively test all potential influences on an index from its own past and other indices. The influences of the other indices can be contemporaneous. The results of graphical modelling are compared to the standard econometric tool for measuring multivariate volatility, the multivariate BEKK-GARCH model. The data used for this investigation is the daily closes of the Standard and Poor's 500 Composite Index (S&P 500), FTSE 100 and Nikkei 225. The period of investigation is from 3 April 2001 to 31 March 2005. The three stock indices are widely followed and over a 24 hour period and there is little overlap in trading hours. The Dickey-Fuller and Phillips-Perron tests confirm that for all three series that the log returns are first order stationary and the index prices are not stationary in mean. JEL CLASSIFICATIONS: C22, C32, C51.
AB - In order to understand volatility transmission between financial assets a multivariate model is essential. This paper looks at using graphical modelling to study volatility transmission. Graphical modelling is a technique that objectively test all potential influences on an index from its own past and other indices. The influences of the other indices can be contemporaneous. The results of graphical modelling are compared to the standard econometric tool for measuring multivariate volatility, the multivariate BEKK-GARCH model. The data used for this investigation is the daily closes of the Standard and Poor's 500 Composite Index (S&P 500), FTSE 100 and Nikkei 225. The period of investigation is from 3 April 2001 to 31 March 2005. The three stock indices are widely followed and over a 24 hour period and there is little overlap in trading hours. The Dickey-Fuller and Phillips-Perron tests confirm that for all three series that the log returns are first order stationary and the index prices are not stationary in mean. JEL CLASSIFICATIONS: C22, C32, C51.
KW - BEKK model
KW - Directed acyclic graph
KW - Spill-over effects
UR - https://www.scopus.com/pages/publications/80052936680
M3 - Conference Publication
AN - SCOPUS:80052936680
SN - 9780975840047
T3 - MODSIM07 - Land, Water and Environmental Management: Integrated Systems for Sustainability, Proceedings
SP - 1399
EP - 1402
BT - MODSIM07 - Land, Water and Environmental Management
T2 - International Congress on Modelling and Simulation - Land, Water and Environmental Management: Integrated Systems for Sustainability, MODSIM07
Y2 - 10 December 2007 through 13 December 2007
ER -